What Salesforce’s Listen Labs Acquisition Means for Insights Researchers

By Luke Williams, Chief Research Officer, Discuss

Salesforce’s agreement to acquire Listen Labs is good news for research. I’m happy for the founders, and I’m pleased to see customer understanding attract this level of attention. For researchers, the question is what we do with that attention.

Technology makes the things we already do faster and more efficient. Eventually, it changes the nature of the way we think about the problem. AI gives research both opportunities. The second deserves much more of our attention.

It is also a large part of why I recently joined Discuss as Chief Research Officer. I saw a platform bringing together quantitative research, qualitative research, and the ability to interpret what we learn across all of it and activate it through an intelligence layer for the purpose of making material decisions. That breadth matters to me as a methodologist. It gives us more ways to investigate a problem and more opportunities to challenge an incomplete answer.

Salesforce’s move puts AI interviewing in the spotlight, and I want researchers to use that attention to think more ambitiously about how we combine methods, develop understanding, and shape the decisions our organizations make.

TL;DR

• Salesforce’s move reinforces the business demand for customer understanding. We should use that attention to raise our ambitions for research.
• I see AI interviewing as an alternative to surveys in some applications. The choice depends on the research objective, the people participating, and the evidence the decision requires.
• High-functioning research teams are multimethod by default. Different methods can illuminate the same problem from different perspectives, including within the same project.
• For consequential investment decisions, I believe multimethod research should be part of the due diligence. We should help leaders test their assumptions before they commit.

Salesforce confirms their belief that research matters.

Salesforce announced the agreement on September 29, describing how Listen Labs will complement Marketing Cloud, Service Cloud, and its broader AI portfolio. That’s a meaningful signal for researchers. Customer understanding is attracting investment from a company whose reach extends far beyond the research department.

I think qualitative research has been underused for years. We are often quick to measure a problem we already recognize and slow to investigate whether we understand it properly. Open-ended, flexible conversations give people room to explain what matters to them, including things we didn’t know to ask about. That can change our understanding of the problem before we decide how to measure it. 

AI interviews could bring that style of inquiry into more of the work organizations already do. That’s worthy of our attention and support.

What did the survey boom teach us?

During my time at Qualtrics, I saw what happened when we helped businesses understand what surveys could do for them. Surveys were already accessible. The opportunity was to translate a method into a purpose. What could it help an organization understand, monitor, or improve? The use cases followed.

As those applications multiplied, the tools reached far beyond specialists. The methodological guidance did not always make the same journey.

As research becomes more accessible and commonplace, we need expertise to scale alongside it. Put capable tools in the hands of experts, and provide methodological support and services where teams need them. Making it possible to run a study is only the beginning. 

In some applications, I see AI interviewing taking on work we previously approached through surveys, with more room for participants to explain their answers conversationally. Follow-up questions alone do not make it unique; survey platforms can support those too. The choice should depend on the research objective and what the design needs to accomplish. 

Human-led interviews give a skilled researcher room to develop trust, investigate an unexpected answer, and adapt the conversation. Surveys and AI interviews can serve different purposes within the same project. None should be treated as a universal substitute for the others. 

Nor does collecting more interviews automatically justify population estimates or a margin of sampling error. Those depend on the sampling and study design, rather than simply on whether a human or AI asked the questions.

The greater opportunity is to combine perspectives into a more comprehensive understanding. Where findings agree, we can examine the strength of that agreement. Where they differ, we have something important to investigate.  

Capacity is great. Attention matters more. 

Much of this conversation focuses on the technology, especially how many interviews it can run and how fast it can analyze them.

I start with the person being asked to participate. Falling response rates should push us to examine what we are asking people to do and whether it feels worth their time. AI interviewing gives us another way to engage them, with more flexibility to participate when it suits them.

Their willingness to give us time and tell us something useful is precious. Making another interview cheap to run doesn’t give them more attention to spend.

My willingness to participate depends on what we are discussing. I may give my wealth manager 20 minutes because the outcome matters deeply to me. After dinner out with my family, I may only be willing to give the restaurant a brief response. The same person can bring very different levels of engagement to different research situations.  

Someone else may be comfortable typing an answer but reluctant to speak on camera. A sensitive conversation may require the trust and permission a skilled interviewer can establish.

Those differences affect who participates and what they are willing to share, so they belong in the research design. AI gives us another way to invite people into research, which means one more design decision to get right.

Multimethod researchers will thrive.

High-functioning research teams are multimethod by default. They know how to choose an approach and how to combine perspectives to develop a more complete understanding.

I expect single-method work to become increasingly concentrated in well-understood applications, such as monitoring a known service problem. When the task is discovery or a consequential investment decision, I believe multimethod research should be the default.

That does not mean adding methods for their own sake. Human-led interviews might illuminate a problem and help shape the hypotheses worth investigating. Depending on the research objective, a survey or AI interviews could help examine those hypotheses across a broader set of participants. Another round of human-led work might help explain a finding that does not make.

There is no mandatory sequence. The research objective determines the design, and the design determines what we can responsibly conclude.

The best researchers understand their work as a craft: a combination of science and art that they learn to apply with judgment. They bring their own strengths to that work, while recognizing where another method or another researcher can contribute something they would otherwise miss.

That is also where interpretation matters. Collecting evidence through several methods is useful only if we can make sense of it together – examining agreement, investigating contradictions, and explaining what it means for the decision.

Discuss is a qualitative and quantitative research platform that combines AI-moderated interviews with human-led research. Researchers can run AI-moderated interviews, live human-led research, online surveys, and computer-assisted telephone interviewing (CATI) in one place. That gives them different ways to reach people and investigate a problem.

My role is to help connect those capabilities to the decisions our customers need to make. Where can AI extend a team’s reach? Where would an experienced interviewer find something we might otherwise miss? What needs to be measured across a broader population before the business acts?

Having the methods available is a starting point. The value comes from knowing how to use them together, where appropriate, and interpreting what their different perspectives reveal.

That’s the work I came here to do: helping research teams take on more ambitious questions and present their evidence with enough conviction that the business acts on it.

Build a research capability you can rely on.

We spend a great deal of time thinking about the impact our research has on the organization. We should periodically give the same scrutiny to the tools, methods, and partnerships that make that work possible.

Can we reach the people we need to understand? Can we change methods when the research objective demands it? Do we have the expertise to design the work well and interpret what comes back? Are we building a capability that can adapt as the organization’s decisions become more complex?

Those questions are central to what I want to help customers address at Discuss.

Our quantitative capabilities help researchers measure and test. Human-led qualitative research provides room for deep exploration. AI interviewing adds another way to gather perspectives at scale. An intelligence layer helps researchers bring what they learn into the decisions the business is making.

The combined value is the ability to select and connect those capabilities around a research objective, with expertise to guide the choices. That is the role I want Discuss to earn with our customers: a strategic partner in building an influential research function.

A tool can help you complete the next study. The relationship should also help you think through what your research team needs to accomplish over the next several years.

Why research needs to move upstream.

Researchers should help shape the question before the organization commissions the study. We need to understand the decision being considered, the assumptions behind it, and what would persuade the business to proceed, stop, or wait.

For your next consequential decision, start there. What is the potential reward? What happens if the assumptions are wrong? Which uncertainty matters enough to investigate, and what combination of methods would give the organization a sound basis for acting?

I believe multimethod research should be part of the due diligence for important investment decisions. An encouraging finding should prompt scrutiny of what supports it, what challenges it, and what remains unknown.

For almost any critical investment decision, I would want both quantitative evidence and a deep understanding of the people behind it. Leaders need to judge the scale of an opportunity and understand why people might behave as the evidence suggests.

Then assess the research by what it changed. Did it expose an assumption? Alter the options? Change the recommendation? Give the organization a sound reason to proceed or to hold back?

Stay close enough to the work to recognize those differences. Conduct some interviews yourself, and listen to the answers.

“You will understand your output better if you understand your input better.”

Then use AI to extend what your team can accomplish.

Salesforce’s move has brought fresh attention to a discipline I care deeply about. We should use that attention to raise our ambitions.

When I joined Discuss, I spoke about permanence of impact. In practice, that means building a lightweight fabric of research instruments that keeps providing useful evidence while we pursue targeted projects to investigate new questions. Each project can strengthen that foundation: revealing a gap, refining a measure, or adding a perspective we were missing. The business gets value continuously, and our capacity to help it grows.

That is what I want to help build at Discuss. Researchers become the architects of an evolving evidence system, helping the organization understand what it knows, where it needs to learn more, and what evidence a consequential decision requires. That is how research becomes infrastructure and how we earn a larger role in shaping the decisions ahead.